AIC★ 51,088⑂ 6,095 forksApache-2.0
JustVugg/colibri
Tiered MoE Inference Engine
Runs large MoE models on commodity hardware by tiering experts: hot on GPU (CUDA/Vulkan), warm in RAM, cold streamed from NVMe.
// KEY FEATURES
- Tiered expert storage across GPU, RAM, and NVMe
- CUDA and Vulkan backends for GPU expert execution
- Zero-dependency C implementation runs on existing hardware
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram